Avoiding Complications in Surgical Resection of Insular Gliomas – Single Surgeon Experience
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.325334
Source of Support: None, Conflict of Interest: None
Keywords: Insular gliomas, outcome, post-operative complications, trans-opercular approachMessage: Transopercular approach is better than transsylvian approach in insular glioma. Most of them are low gradehence aggressive and safe excision should be done.
Gliomas in the insular region account for one-fourth of all low grade gliomas (LGG) and nearly 10% of all high grade gliomas (HGG). More often, these gliomas are surrounded by microvasculature supplying language and motor function areas. In fact, insula is implicated in cognitive functions like memory, gustation, and olfaction. Therefore, insular gliomas remain one of most challenging locations for aggressive resection., The natural history of insular gliomas is characteristically unpredictable with a prolonged and slow progressive course; generally less aggressive than tumors in other locations. Sanai et al., in 2010, had proposed a quadrant style classification (Zone I – IV) of the location of the glioma in the insula. In the same study, the authors reported that aggressive resection of insular gliomas of all grades can be accomplished with an acceptable morbidity profile. Furthermore, the authors quote that extent of resection (EOR) is predictive of improved overall survival and progression free survival. However, surgical resection also carries substantial complication rates, similar to unselected craniotomy. Recently, Benet et al., in a “cadaveric” study reported that trans-cortical/trans-opercular (TO) approach to the insula provides better exposure to insula and surgical freedom than the trans-sylvian (TS) approach. In the current study, a retrospective analysis of our experience with surgical resection in patients with low grade insular gliomas, we report the strategies we employed to avoid complications immediately after surgery and at three months.
Retrospective analysis of data in 61 consecutive patients with insular gliomas undergoing surgical resection between May 2013 and May 2016 was done. All surgical resections were performed by the senior author (M.K.P.). Histopathology review was performed based on WHO guidelines. Patients with WHO grade I gliomas were excluded from analysis because their biology and natural history differ substantially from Grade II–IV lesions, and their inclusion could bias the analysis toward improved survival with a more extensive resection. Clinical data were collected from patient records and telephone interviews. The study was approved by Institutional ethics committee. An informed consent to contact patients on telephone was obtained from all the participants or their caregivers. In 40 out of 61 patients, surgery was planned (decision of biopsy or resection) after detailed study of fMRI and DTI tractography. The protocol of fMRI and DTI have been previously reported. All patients underwent surgical resection under general anesthesia. All patients were recruited after obtaining an informed consent. The study was approved by institutional ethics committee (approval number: No - KFRC/SRS/2016/59-02; Date - 22/12/2016).
Since it is known that surgical outcome is not significantly different between awake craniotomies and resections done under general anaesthesia; surgery under general anesthesia was preferred to avoid excessive pain to the patient caused by awake craniotomy. The anatomical land marks used to limit resection were perinsular sulcus (circular sulcus) limen insulae, and the lenticulostriate arteries. Insular position of the tumor was termed based on these anatomical land marks. In cases where there was involvement of the eloquent area by the tumor or corticospinal tract, surgical procedure was tailored based on anatomical land marks, using sub-cortical stimulation of the corticospinal tracts and intra operative neuro-navigation. The relative position of the glioma with respect to lenticulostriate arteries was evaluated using intra-operative angiography, using indocyanine green (ICG). Since it remains unclear which extent of resection (EOR) cut-off provides the optimal benefit, EOR was defined as resection extensive enough but relatively safe (presumed >90%).
In the TS approach, the sylvian fissure was split and retraction was done over the opercula to expose the insula. In this approach, the superficial and deep sylvian cisterns are opened widely throughout the fissure. The first 38 consecutive patients underwent resection using the TS approach, as it was the standard at the study centre till early 2016. Subsequent 23 patients underwent surgical resection using the TO approach. In the TO approach, different degrees of cortical resection were used to expose the insula. In the TO, the bridging veins were preserved and the space to access the insula resulted from sub-pial resection below middle cerebral arteries and the superficial sylvian veins.
Berger-Sanai Classification of the insular glioma was done as recommended by Sanai et al. Insula was divided into four zones, and tumor location was assigned to one or more of these zones. Along the horizontal plane in a sagittal view, the insula was bisected along the sylvian fissure. A perpendicular plane intersected at the level of the foramen of Monro. The resultant anterior-superior, posterior- superior, posterior-inferior, and anterior-inferior quad-rants were designated Zones I, II, III, and IV, respectively. In cases where the tumor occupied all four zones, the glioma were defined as “giant.”
Patient outcome measurements
The primary outcome measures were neurological complications up to 3 months after the day of discharge or death in the immediate post-operative period. Outpatient follow-up were done at one week, 4-6 weeks, and 3 months after surgery. Post-operatively and at each follow-up appointment, patients underwent neurological examination by the senior neurosurgeon or a neurosurgery resident. Neurological complication was defined as new-onset deficits or worsening of existing deficits related to motor, sensory, visual, and language function. Follow-up MRI was done at three months of surgery.
All continuous variables are expressed as mean ± SD; whereas categorical variables are reported as frequency or percentages. The study population was divided into TO (n = 23) and TS (n = 38) groups. Differences between the groups for continuous variables were analyzed using independent student-tests and Chi-square test was used to analyze categorical variables. A P value ≤0.05 was considered significant. Whereas, a P < 0.1 is reported as a trend. All statistical analysis was done using Statistical Package for social Sciences (SPSS) for windows, version 17.0, IBM Computers, New York, USA.
The average age of the study population was 42.57 ± 10.98 years with 41 (67.2%) men. Glioma was on the right side in 35 (57.3%) patients. Surgery for recurrent glioma was performed in three (4.9%) patients. Pre-operatively, the symptoms most often reported were seizures of recent onset in 34 (55.7%) followed by headache in 22 (36.1%) patients. On clinical examination, memory deficits were observed in two (3.3%), speech deficits and motor deficits in seven (11.5%) patients each. Clinical signs of raised intra cranial pressure (ICP) were observed in 11 (18.0%) patients. The average MIB index of the entire group was 10.1 ± 13.9. Pre-operatively, motor deficits were more often observed in patients undergoing TS approach (0.0% vs 18.4%; P = 0.038), [Table 1].
Twenty three (37.7%) patients underwent the TO approach, while the remaining thirty eight (62.3%) underwent TS approach. Among the 23 patients who underwent resection using TO approach, gliomas was on the left side in 10 patients. Among these 10 patients, the glioma was distant from language area in five patients, abutting language area in one patient; whereas fMRI was not done in the remaining 4. In the immediate post-operative period, a higher number of patients under TS approach had post-surgical complications (8.6% vs 34.2%; P = 0.032). There was no difference between the surgical approaches for number of patients undergoing extent of resection >90%. However, the outcome at 3 months did not significantly differ between the surgical approaches [Table 2].
On histopathological examination, the most common WHO grade of tumor was Grade II in 42 (68.8%) patients, followed by grade IV in 12 (19.7%) patients and Grade III in seven (11.5%) patients. There were no differences between the surgical approaches for WHO grade of tumors [Table 3]. Oligo-astrocytomas were the most common types of gliomas observed in histo-pathological examination, observed in 14 (23.0%) patients, followed by fibrillary/diffuse astrocytomas in 13 (21.3%), and glioblastomas in 12 (19.7%). There were no differences between surgical approach for both WHO grade and histopathological findings, [Table 3].
On the Berger-Sanai classification, the most common location for the glioma was Zone I, observed in 19 (31.1%) patients, followed by Zone I + IV in 13 (21.3%) patients. While the seven patients had glioma in Zone IV (11.5%), in six patients (9.8%), the glioma was located in Zone III. Giant gliomas and gliomas located in Zone II were observed in four (6.5%) patients each. Three patients (4.9%) each had glioma located either in Zones II + III or zones III + IV. The remaining two (3.2%) patients had glioma in zone I + II.
Among the 19 patients with glioma in Zone I, 9 (47.3%) were operated through TO approach. All 4 patients with glioma exclusive to Zone II underwent surgery through TS approach. Among the 6 patients with glioma in Zone III, 5 were operated though TS approach. Whereas, in the 7 patients with glioma exclusive to Zone IV, 4 patients were operated though TS approach, [Figure 1].
Among the patients with glioma extended more than one zone, in the four patients with giant glioma, two patients each underwent TS and TO approach respectively. All three patients with glioma in zones II and III underwent TS approach, [Figure 2]. When analyzed for complications between Berger Sanai Zones versus approach to surgery, both the surgical approaches did not differ significantly for the number of complications, [Table 4].
In the current study on patients who underwent surgical resection of insular gliomas, we report that a utilization of both Berger-Sinai classification and selecting of either trans-opercular (TO) approach or trans-sagittal (TS) approach is associated with lower complications in the immediate post-operative period. Trans-opercular is associated with lower number of complications in the immediate post-operative period than Trans-sylvian approach. We have previously reported that use of fMRI and DTI tractography improves outcome and reduces complications during immediate post-operative period in patients undergoing surgical resection of insular gliomas.
Daffau et al. reported complications in immediate post-operative period in 59% of the patients exclusively with WHO grade II gliomas, who underwent surgical resection. Similarly, we report significantly lower complications (47.5%) in the current series that included patients with WHO grades II, III and IV. The reported incidence of complications is higher than the 36% reported by Lang et al., since the authors included only speech and motor deficits as complications, in the current study we also included, language, visual, sensory deficits along with death as complication. Moreover, if only motor and speech deficits were considered, only 22.9% patients had deficits; lower than those reported by Lang et al. Moreover, our findings are similar to Shankar A et al., who followed patients with low-grade insular astrocytomas (WHO grade II), following stereotactic biopsy and radiotherapy for more than 24 months. The authors performed complete memory and language function testing 11 patients; at follow-up, six patients showed a slight improvement, while the score worsened in 3 and remained the same in 2 patients.
Additionally, we report that TO approach is associated with lower number of complications in immediate post-operative period. Sanai et al., based on the Berger-Sanai classification system, showed that “it is feasible to aggressively resect gliomas of all grades in the insula”. The findings of the current series reinforce that aggressive resection based on the Berger Sanai classification in fact reduces morbidity in the immediate post-operative period and also reduces mortality at one year.
The mortality rate at nearly 9.8% in the current study is lesser than existing literature, is probably because of utilization of both Berger-Sanai classification and TO approach. Importantly, none of the 6 deaths were neither related to the surgical procedure, nor observed in the immediate post-operative period. Although all the observed deaths were in patients who underwent TS approach, our findings are not conclusive. Perhaps a larger population and further research is required to compare both the approaches.
Benet et al. in a cadaveric study of surgical approach to insular gliomas, have recently showed that the TO corridor provides the greatest access to lesions. The authors further state that TO approach provides the best overall surgical exposure, window, and freedom. Our findings add that, perhaps because of this feasibility of overall surgical exposure, more numbers of patients undergo more than 90% of resection of the glioma at a lesser number of post-surgical complications.
A better exposure of the gliomas margins perhaps avoids unnecessary manipulation of the vessels and resultant vaso-spasm. This probably reduces the incidence of complications in the immediate post-operative period. Lang et al., previously reported that “speech and motor dysfunction appeared to result most often from manipulation of the middle cerebral artery (MCA), interruption of the lateral lenticulostriate arteries (LLAs), interruption of the long perforating vessels of the second segment of the MCA (M2), or violation of the corona radiata at the superior aspect of the tumor”. Moreover, Shah et al., recently reported that intra-operative use of ICG helps to identify and preserve the long perforating branches of the MCA that course through the tumor and traverse onward to supply the corona radiata. Perhaps, the use of ICG in the current study may have contributed to lower post-operative complications.
Benet et al. further suggest that, “if the tumor is limited at the sylvian line exclusively either in Zone I or Zone IV, tumor can be sufficiently exposed through TS approach. In the current study, 19 patients had glioma exclusive to Zone I and 10 of them underwent resection through TS approach. Similarly, of the sev7en patients with glioma in zone IV, 4 of them underwent resection through TS approach. It is commonly accepted that TS approach requires large amounts of retraction to expose the insula.,,, Perhaps, because of this retraction required, the complication in TS approach we observed both in Zone I and Zone IV are significantly higher. In the current study, the maximal retraction length that we applied to each operculum without damaging the venous system is two cms, in agreement with previous studies. Current literature suggests that TO approaches to sub-cortical lesions are safe.,, Furthermore, TO approach provides a better surgical outcome than the other options.
Daffau et al., previously proposed the systematic use of intraoperative subcortical electrical stimulation, as an easy, safe, accurate, and reliable method of direct detection of the internal capsule. Moreover, in the patients where the authors resected the primary motor face area of the nondominant hemisphere, cortical-subcortical stimulation allowed to differentiate the motor face structures from the motor hand regions, which it is imperative to respect. The authors conclude that “with brain mapping methods, it becomes possible to perform glioma surgery using functional and not anatomic boundaries in the insular location”. Importantly, use of intra-operative sub-cortical stimulation reduced post-operative motor complications while assuring optimal resection of the glioma. Utilizing, sub-cortical electrical stimulation in the current study might have helped to reduce complications and improve outcome. However, our study did not have enough statistical power to evaluate role of intra-operative sub-cortical stimulation either individually or in combination of the other reported factors.
Strengths and limitations: The current study is a single surgeon experience and reflects findings from a consistent environment and patient care protocols, with little scope for variability that is often attributed to different models of care at different institutions in multi-centre studies. We included multiple comparisons involving distribution of gliomas into single or multiple zones with a potential for inherent errors, however the findings seem to be clinically plausible. Despite these limitations, our study adds relevant information about reducing complications in surgical resection of insular gliomas.
In patients undergoing surgical resection of insular gliomas, we report that morbidity and mortality in immediate post-operative period can be reduced by a) pre-surgical assessment of confinement of glioma in respect to lenticulo-striate arteries. b) Intra-operative use of functional-MRI, DTI tractography and ICG angiography. c) Application of Berger-Sinai classification to localize the glioma d) selecting either TS or TO approach based on Berger-Sinai classification. Contribution of each of the above mentioned four factors either individually or in combination could be studied in detail in larger study population.
Financial support and sponsorship
Conflict of interest
None of the authors have any conflict of interest to declare.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]